Abstract During fracture stimulation treatments, particle-laden slurries may damage pumping equipment and have been shown to erode perforation tunnels. During subsequent flowback operations, minor to severe erosion of surface equipment may be observed if formation sand and/or proppants are produced. This paper compares the erosivity of sand-based and ceramic proppants. This paper will examine the erosion mechanisms and demonstrate that theoretical and laboratory measurements are in agreement that low density ceramic proppants, due to the superior roundness and sphericity, are less erosive than more angular sand products. Real-world results from the oilfield, the waterjet industry, pump and valve manufacturers, and the abrasive industry will be reviewed to substantiate the conclusion that proppant angularity is a dominant parameter that increases erosion in typical pumping and flowback conditions. Impingement tests designed to simulate flowback conditions demonstrate that substituting a spherical lightweight ceramic may reduce wellhead or other impingement surface erosion by up to 95% compared to an angular sand. The highest quality frac sand was nine times more erosive than lightweight ceramic (LWC) in impingement studies. Additional studies utilizing intentionally crushed proppant to increase angularity demonstrate that proppant shape is a dominant factor which controls erosion. Surprisingly, flowback testing has shown that resin coatings applied to proppant have the potential to increase erosion by 15-fold. The authors believe this is partially due to a focusing effect attributed to electrostatic charge. Test alterations with multiphase flow have reduced, but not eliminated, this effect. Slurry abrasion testing has shown that the highest quality frac sands are over 250% as abrasive as LWC. Although frac sands are often specified to withstand several thousand pounds of closure stress, significant sand degradation was observed after exposure to a reciprocating five pound weight. Erosivity and abrasivity of all proppant types were shown to increase with proppant damage. Erosion within chokes was found to be three times higher with sand than with lightweight ceramic. Despite conventional wisdom to the contrary, a number of recent laboratory studies have demonstrated that spherical proppants are less likely to flow back from the fracture as compared to more angular products. All data indicate that equipment erosion occurring during the stimulation treatment and subsequent flowback can be substantially reduced through the use of spherical lightweight ceramic proppants.